ArticlePDF Available

Prospective predictors of body dissatisfaction in adolescent girls and boys: A five-year longitudinal study

Authors:

Abstract

This study investigated prospective risk factors for increases in body dissatisfaction in adolescent girls and boys in the Eating Among Teens Project. At the time of first assessment (Time 1), participants were a cohort of early adolescent girls (N=440) and boys (N=366) and a cohort of middle adolescent girls (N=946) and boys (N=764). Participants were followed up 5 years later (Time 2). Potential prospective risk factors examined included body mass index, socioeconomic status, ethnicity, parent dieting environment, peer environment, and psychological factors. Predictors of Time 2 body dissatisfaction were Time 1 body dissatisfaction, body mass index, socioeconomic status, being African American, friend dieting and teasing, self-esteem, and depression. However, the profile of predictors differed across the samples.
Prospective Predictors of Body Dissatisfaction in Adolescent Girls and
Boys: A Five-Year Longitudinal Study
Susan J. Paxton
La Trobe University
Marla E. Eisenberg and Dianne Neumark-Sztainer
University of Minnesota, Twin Cities Campus
This study investigated prospective risk factors for increases in body dissatisfaction in adolescent girls
and boys in the Eating Among Teens Project. At the time of first assessment (Time 1), participants were
a cohort of early adolescent girls (N 440) and boys (N 366) and a cohort of middle adolescent girls
(N 946) and boys (N 764). Participants were followed up 5 years later (Time 2). Potential
prospective risk factors examined included body mass index, socioeconomic status, ethnicity, parent
dieting environment, peer environment, and psychological factors. Predictors of Time 2 body dissatis-
faction were Time 1 body dissatisfaction, body mass index, socioeconomic status, being African
American, friend dieting and teasing, self-esteem, and depression. However, the profile of predictors
differed across the samples.
Keywords: risk factors, body dissatisfaction, longitudinal, adolescent girls and boys
Prospective risk factors for the development of body dissatis-
faction during adolescence have received increasing attention as
the prevalence and negative consequences of body dissatisfaction
have become more widely recognized. In recent large community
samples, body dissatisfaction has been reported by 24%– 46% of
adolescent girls and 12%–26% of adolescent boys (Neumark-
Sztainer, Story, Hannan, Perry, & Irving, 2002; Presnell, Bearman,
& Stice, 2004; Stice & Whitenton, 2002). Ricciardelli and McCabe
(2001) found only 12.0% of girls and 16.6% of boys were their
desired size. Indicative of more severe weight concern, in one
study, 12.4% and 4.6% of girls and boys, respectively, reported
using extreme weight loss strategies (e.g., fasting, use of diet pills,
laxative abuse, or vomiting; Neumark-Sztainer, Story, Hannan,
Perry, & Irving, 2002), and in another, 12.4% of boys reported
frequently using or thinking about using food supplements or
steroids to gain muscle (Ricciardelli & McCabe, 2003). A meta-
analysis has demonstrated body dissatisfaction to be “one of the
most consistent and robust risk and maintenance factors for eating
pathology” (Stice, 2002, p. 833). Further, body dissatisfaction has
been found to prospectively predict development of depressed
mood (Holsen, Kraft, & Roysamb, 2001; Stice & Bearman, 2001).
Given the widespread distress and negative outcomes associated
with body dissatisfaction, it is important to understand factors that
predict its development.
The major theories of the development of body dissatisfaction
suggest that its etiology is multifactorial. Researchers propose that
physical characteristics—especially larger body size and cultural–
social environments emphasizing thinness and appearance—and
psychological factors—including low self-esteem, depression, and
beliefs about the importance of thinness—increase risk of devel-
opment of body dissatisfaction (e.g., Cash, 2002; Paxton, Norris,
Wertheim, Durkin, & Anderson, 2005; Stice, 1994; van den Berg,
Thompson, Obremski-Brandon, & Coovert, 2002). However, these
models and related empirical studies have not given great consid-
eration to potential differences in risk factors between girls and
boys, despite social and appearance pressures varying considerably
between girls and boys (e.g., Cafri et al., 2005; Cohane & Pope,
2001; Wichstrom, 1999). In addition, it has largely been assumed
that risk factors for increases in body dissatisfaction will be similar
across adolescence despite developmental tasks and roles changing
markedly from early to late adolescence. Consequently, in the
present research, we aimed to identify psychosocial predictors of
increases in body dissatisfaction over a 5-year follow-up period,
using a multivariate approach, in both girls and boys, at two
developmental stages, from early to middle adolescence and from
middle to late adolescence.
Cross-sectional research has supported potential predictive roles
for increases in body dissatisfaction for a range of psychosocial
factors, including body size, ethnicity, peer and parent environ-
ments that emphasize the importance of thinness, and personality
variables. In the modernized world in which slimness is valued, it
has been proposed that a large body size, representing a departure
from the current beauty norm, increases vulnerability to body
dissatisfaction (e.g., Stice, 1994, 2002; Thompson, Heinberg,
Altabe, & Tantleff-Dunn, 1999; Wertheim, Paxton, & Blaney,
2004). Consistent with this view, body mass index (BMI) is the
most reliable correlate of body dissatisfaction (e.g., McCabe &
Ricciardelli, 2003). Within cultures, however, there appear to be
variations in beauty norms among ethnic groups, and, in particular,
it has been suggested that the ideal figure in females is larger in
Susan J. Paxton, School of Psychological Science, La Trobe University,
Melbourne, Australia; Marla E. Eisenberg, Division of General Pediatrics
and Adolescent Health, Department of Pediatrics, University of Minnesota,
Twin Cities Campus; Dianne Neumark-Sztainer, Division of Epidemiology
and Community Health, School of Public Health, University of Minnesota,
Twin Cities Campus.
This research was supported by Maternal and Child Health Bureau (Title
V, Social Security Act) Health and Human Services Grant R40 MC 00319
(Dianne Neumark-Sztainer, principal investigator).
Correspondence concerning this article should be addressed to Susan J.
Paxton, School of Psychological Science, La Trobe University, Melbourne
3086, Australia. E-mail: susan.paxton@latrobe.edu.au
Developmental Psychology Copyright 2006 by the American Psychological Association
2006, Vol. 42, No. 5, 888 899 0012-1649/06/$12.00 DOI: 10.1037/0012-1649.42.5.888
888
African American than in other ethnic subcultures (Halpern, Udry,
Campbell, & Suchindran, 1999; Neumark-Sztainer, Croll, et al.,
2002). In addition, the majority of studies (e.g., Akan & Grillo,
1995; Kelly, Wall, Eisenberg, Story, & Neumark-Sztainer, 2005;
Neumark-Sztainer, Croll, et al., 2002; Neumark-Sztainer, Story,
Faibisch, Ohlson, & Adamiak, 1999), but not all (e.g., Striegel-
Moore et al., 2000), have found that African American girls are
less likely than European American girls to express body dissat-
isfaction. Similarly, African American boys are more likely to
have high body satisfaction compared with other ethnic groups
(Neumark-Sztainer, Croll, et al., 2002). Norms may also vary
among socioeconomic groups independent of ethnicity and BMI.
However, such potential differences have not been frequently
examined.
It has also been proposed that immediate social environments
that value appearance or slimness contribute to greater value being
placed on appearance by the child. As this ideal is almost invari-
ably difficult to achieve, higher internalization of sociocultural
ideals generates greater body dissatisfaction (e.g., Stice, 1994; van
den Berg et al., 2002). Family attitudes toward weight and shape
reflected in parental weight loss practices and encouragement of
children to diet may model or reinforce appearance values that are
risk factors for body dissatisfaction. Cross-sectional research pro-
vides support for relationships between encouragement to diet by
parents and body dissatisfaction and dieting in adolescents but
provides less consistent support in relation to modeling effects
(Fulkerson et al., 2002; Keery, Eisenberg, Boutelle, Neumark-
Sztainer, & Story, in press; Wertheim, Martin, Prior, Sanson, &
Smart, 2002; Wertheim, Mee, & Paxton, 1999). Parental child-
feeding style has also been implicated in childhood eating and
obesity, which may indirectly affect body image. In particular,
parental restriction of and control over child food intake may
increase desire to obtain restricted food, eating in the absence of
hunger, and obesity (Birch, Fisher, & Davison, 2003; Fisher &
Birch, 1999, 2002; Spruijt-Metz, Lindquist, Birch, Fisher, &
Goran, 2002), although these relationships have not been consis-
tently observed (Robinson, Kiernan, Matheson, & Haydel, 2001).
Similarly, peer environments that place high value on slimness,
reflected in friend dieting behaviors and weight and shape teasing,
have been proposed as risk factors for body dissatisfaction. Qual-
itative research has indicated that peer discussions of appearance
may result in insecurity and body image concerns (Nichter, 2000;
Wertheim, Paxton, Schutz, & Muir, 1997). Quantitative cross-
sectional studies have also consistently reported associations be-
tween body dissatisfaction and peer conversations about appear-
ance, perceived importance placed on appearance by friends, and
appearance teasing (Barr Taylor et al., 1998; Eisenberg, Neumark-
Sztainer, & Story, 2003; Jones, Vigfusdottir, & Lee, 2004; Paxton,
Schutz, Wertheim, & Muir, 1999; Wertheim, Koerner, & Paxton,
2001). In 10th-grade girls, Paxton and colleagues (1999) found
that perceived friend concern about thinness, friend talk about
dieting, peer pressure to be thin, and weight teasing predicted body
dissatisfaction after BMI and psychological and family variables
had been taken into account. Similarly, Jones and colleagues
(2004) observed significant correlations between appearance con-
versations and body dissatisfaction, and peer appearance criticism
(teasing) and body dissatisfaction in both adolescent girls and
boys. Eisenberg and colleagues (2003) reported that not only is
weight teasing associated with body dissatisfaction in adolescent
girls and boys, but it is also correlated with low self-esteem and
high depressive symptoms even after controlling for BMI.
In a cross-sectional study, combined perceived family and peer
weight norms were associated with body concerns in boys and
girls (Neumark-Sztainer, Wall, Story, & Perry, 2003). In addition
to the importance placed on thinness by peers and family, within
an individual, internalization of social appearance ideals, thinness
in girls, and muscularity in boys are correlated with body dissat-
isfaction (Jones et al., 2004).
Finally, personality attributes (self-esteem and perfectionism)
and related psychological functioning (especially depression) have
been proposed to be associated with a negative and self-critical
view of the self generally and of the body in particular, contrib-
uting to the development of body dissatisfaction (e.g., Cash, 2002;
Stice, 1994; van den Berg et al., 2002). Cross-sectional support for
relationships among these variables has also been observed
(Ohring, Graber, & Brooks-Gunn, 2002; Paxton et al., 1999; Wade
& Lowes, 2002).
Prospective Research
Although theoretical models and cross-sectional studies may
suggest a risk factor role for different variables, longitudinal or
experimental research is required to confirm the temporal prece-
dence of a potential risk factor. Numerous longitudinal studies
have explored putative prospective risk factors for body dissatis-
faction or weight concerns in adolescent girls, whereas fewer have
examined risk factors in boys (Wertheim et al., 2004). These
studies have included consideration of more proximal risk factors,
with follow-up periods of usually 1–2 years, whereas a smaller
number have examined more distal risk factors, with follow-up
periods of 3– 8 years.
After researchers controlled for baseline body dissatisfaction,
higher baseline BMI or change in BMI was the most reliable
predictor of increases in body dissatisfaction in girls and boys in
both short- and long-term follow-up studies (e.g., Field et al.,
2001; Halpern et al., 1999; Ohring et al., 2002). However, in a
1-year follow-up of 17-year-olds, Presnell and colleagues (2004)
found this relationship disappeared when negative affect was also
considered in the analysis. The influence of norms held by differ-
ent ethnic and socioeconomic groups has not received much at-
tention in longitudinal studies. However, Halpern et al. (1999), in
a study of girls, observed that being European American compared
with African American predicted increases in body dissatisfaction
over 2 years.
As attention has turned to the potential role of immediate social
environments in the development of body dissatisfaction, more
recent longitudinal analyses have included assessments of family
and peer influences as predictor variables. To date, studies of this
kind have had relatively short-term follow-up periods (1–2 years).
With one exception (Byely, Archibald, Graber, & Brooks-Gunn,
2000), the studies offer support for the predictive role of social
factors on change in body dissatisfaction. Field et al. (2001)
observed more frequent development of weight concerns among
girls when thinness was perceived as important to either parent,
and among boys when their mother was trying to lose weight. Stice
and Whitenton (2002) found perceived pressure to be thin (from
parents, peers, and media) predicted body dissatisfaction in girls,
whereas Presnell et al. (2004) found perceived pressure to be thin
889
BODY DISSATISFACTION IN ADOLESCENT GIRLS AND BOYS
from peers predicted body dissatisfaction in a univariate analysis,
although it did not predict body dissatisfaction when considered in
a multivariate analysis. In an interesting study by Jones (2004), the
role of Time 1 peer appearance conversations on change in body
dissatisfaction was mediated by body comparison tendency in
girls. However, in boys, peer appearance conversations and con-
cerns did not play a strong role. Rather, internalized appearance
ideals predicted change in body dissatisfaction. In a longer
follow-up (3 years), Cattarin and Thompson (1994) found weight
teasing predicted body dissatisfaction. Notably, thin-ideal internal-
ization and dysfunctional beliefs about appearance have also been
found to be proximal predictors of increases in body dissatisfaction
(Spangler, 2002; Stice, 2001; Stice & Whitenton, 2002).
There is mixed support for a risk factor role for personality and
psychological well-being factors after controlling for initial body
dissatisfaction. In studies of relatively short-term follow-up,
Presnell et al. (2004) observed negative affect predicted increases
in body dissatisfaction in boys but not in girls, whereas Stice and
Whitenton (2002) did not observe a relationship between depres-
sion and body dissatisfaction in girls. In studies with long-term
follow-up periods, Ohring et al. (2002) found poor emotional
control in girls predicted increases in body dissatisfaction, and
Martin and colleagues (2000) found negative emotionality in boys
predicted later body dissatisfaction, although they did not control
for baseline body dissatisfaction. On the other hand, Holsen and
colleagues (2001), after controlling for Time 1 body dissatisfac-
tion, did not find depression prospectively predicted body dissat-
isfaction in girls or boys. Despite a consistently strong cross-
sectional relationship between body dissatisfaction and low self-
esteem, few studies have examined the role of self-esteem as a
prospective risk factor for increases in body dissatisfaction during
adolescence after controlling for baseline body dissatisfaction.
Research Aims
Cash (2002) has theorized that causes of body dissatisfaction are
likely to be both historical (e.g., BMI, teasing history, and self-
esteem) and proximal (everyday experiences that activate negative
body image schema). Within this context, it is likely that different
risk factors might be identified using different follow-up periods.
In particular, longer follow-up periods are more likely to detect
more historical factors and those relatively stable within an indi-
vidual or environment, whereas shorter follow-up periods may
detect specific or precipitating risk factors. Only three studies that
control for baseline body dissatisfaction, however, have used long-
term follow-up periods of 3– 8 years (Cattarin & Thompson, 1994;
Holsen et al., 2001; Ohring et al., 2002). In addition, only one has
studied boys (Holsen et al., 2001), and as physical and psycholog-
ical development and social and appearance pressures differ be-
tween girls and boys, and as short-term follow-up studies have
found differences in predictor variables for females and males,
there may also be differences between girls and boys in long-term
predictors. Further, none of the studies with a longer term
follow-up have simultaneously examined BMI, sociodemographic
factors, family and peer weight and shape environments, and
psychological variables as observed in shorter follow-up studies
(e.g., Presnell et al., 2004), and examination of univariate effects
alone may provide a misleading picture.
Finally, there are marked differences in psychosocial develop-
ment, role expectations, and social pressures during early to mid-
dle adolescence and during middle to late adolescence that may
result in different patterns of prospective risk factors for body
dissatisfaction in these different developmental stages. For exam-
ple, during early years of adolescence, most girls and boys expe-
rience marked physical changes associated with puberty (Tanner,
1989). In addition, typically during these years while the family
environment remains very important, there is a strong peer orien-
tation and high importance placed on peer acceptance (Berndt &
Hestenes, 1996). During later years of adolescence and young
adulthood, however, typically there are continued but fewer phys-
ical changes (Whitbourne, 2002), greater interest in heterosexual
relationships (Levesque, 1993), greater independence, and the
challenges of moving into adult environments in tertiary education
and the workplace (Whitbourne, 2002). In light of such marked
developmental differences, we aimed to explore potential differ-
ences in predictors of increases in body dissatisfaction in these
different phases.
In summary, the present study extended existing research into
prospective risk factors for body dissatisfaction by using a rela-
tively long follow-up time of 5 years; examining a range of
physical, sociocultural, and psychological variables simulta-
neously as potential risk factors for increases in body dissatisfac-
tion; studying both girls and boys; and including youth at two
different developmental stages. Cohorts of early and middle ado-
lescents, each containing girls and boys, involved in the Eating
Among Teens Project (Project EAT-I) were reassessed after 5
years (Project EAT-II). The central aim of the study was to
examine whether psychosocial variables (parent dieting environ-
ment, friend dieting and peer teasing, self-esteem, and depression)
would predict increases in body dissatisfaction after initial body
dissatisfaction, BMI (the most consistent predictor of increased
body dissatisfaction), and demographic variables (socioeconomic
status [SES] and ethnicity) were taken into account. A further aim
was to examine predictors of increases in body dissatisfaction in
early and middle adolescent girls and boys separately, so that
different patterns of predictors could be identified in these groups.
Guided by previous research, we hypothesized that BMI would
predict increases in body dissatisfaction and being African Amer-
ican would protect against increases in body dissatisfaction in all
groups. In addition, we hypothesized that friend dieting and peer
teasing would predict increases in body dissatisfaction in female
samples, whereas depression would predict increases in body
dissatisfaction in male samples. As existing theory and research
provided little guidance about potential differences in risk factor
roles of parent dieting environment and self-esteem in girls com-
pared with boys at different developmental stages, no specific
hypotheses were made in relation to these variables.
Method
Participants
Project EAT-I was established as a large epidemiological study of
socioenvironmental, personal, and behavioral determinants of dietary in-
take and weight status among adolescent girls and boys from diverse ethnic
and socioeconomic backgrounds (Neumark-Sztainer, Story, Hannan, &
Croll, 2002). Junior and senior high school students in 31 Minnesota
schools (N 4,746) completed in-class surveys and anthropometric mea-
890
PAXTON, EISENBERG, AND NEUMARK-SZTAINER
sures during the 1998 –1999 academic year. Project EAT-II aimed to
resurvey all original participants 5 years later (2003–2004) as the early
adolescent cohort progressed from early adolescence (junior high school)
to middle adolescence (high school), while the middle adolescent cohort
progressed from middle adolescence (high school) to late adolescence
(post-high school).
For Project EAT-II, surveys were mailed to the address provided by
participants during Project EAT-I. If the survey was not returned, subse-
quent mailings included two reminder postcards, an additional survey, a
second survey sent by Federal Express, and a final survey sent in a
neon-colored envelope with a mechanical pencil with which to complete
the survey. If mail was returned because of an incorrect address, Internet
tracking services were used to identify a current address where possible,
and the mailing procedure recommenced. The University of Minnesota’s
Institutional Review Board Human Subjects Committee approved all
protocols.
Following these procedures, 1,074 (22.6%) of the Project EAT-I sample
were lost to follow-up for several reasons, including missing contact
information at Project EAT-I (N 411) and no address found at follow-up
(N 591). Of the remaining 3,672 participants contacted by mail, 2,516
completed surveys, representing 52.0% of the original cohort and 68.4% of
participants who were contactable for Project EAT-II. The final sample of
adolescents who completed valid surveys in both Project EAT-I (Time 1)
and Project EAT-II (Time 2) consisted of 440 girls in the early adolescent
cohort (Time 1 mean age 12.7 years, SD 0.74), 366 boys in the early
adolescent cohort (Time 1 mean age 12.8 years, SD 0.76), 946 girls
in the middle adolescent cohort (Time 1 mean age 15.8 years, SD
0.81), and 764 boys in the middle adolescent cohort (Time 1 mean age
15.9 years, SD 0.78). The ethnic–racial background of the sample was
61.9% European American, 11.1% African American, 4.5% Hispanic,
17.8% Asian, 1.9% Native American, and 2.7% of mixed or other ethnic-
ity. There was also diversity in SES: 13% low, 17% middle–low, 25%
middle, 28% middle– high, and 17% high SES.
Measures
Body dissatisfaction. Time 1 and Time 2 body dissatisfaction were
assessed with a modified version of the Body Shape Satisfaction Scale
(Pingitore, Spring, & Garfield, 1997). Each respondent rated his or her
satisfaction with 10 aspects of his or her body shape and parts (e.g., height,
weight, body shape, waist, body build, and shoulders) on 5-point Likert
scales ranging from 1 (very satisfied)to5(very dissatisfied). Responses on
each item were summed. Cronbach’s alpha at Time 1 was .92 and at Time
2 was .93. Higher scores indicated greater dissatisfaction.
Weight status. Height and weight were measured using standardized
equipment and procedures used to calculate BMI (weight in kilograms
divided by the square of height in meters). Because 11% of girls and 10%
of boys were missing observed height and weight, self-reported height and
weight were substituted in these cases to reduce the missing percentage to
3% of girls and 2.5% of boys, as the self-reported and observed measures
were highly correlated (girls: r .85, p .001; boys: r .88, p .001).
Demographic variables. Gender, age, ethnicity, and SES were based
on Time 1 self-report. The original seven-category ethnicity variable was
collapsed into three categories of European American, African American,
and other–mixed ethnicity for use in analyses. The prime determinant of
SES was parental education level, defined by the highest level of educa-
tional attainment of either parent. Other variables taken into account in
assessing family SES included family eligibility for public assistance,
eligibility for free or reduced-cost school meals, and employment status of
the mother and father (Neumark-Sztainer, Story, Hannan, & Croll, 2002).
Parent dieting environment. A scale assessing parent dieting environ-
ment was constructed with four items. The following statements were rated
on 4-point scales ranging from 1 (not at all)to4(very much): “My
mother/father diets to lose weight or keep from gaining weight” (two
items) and “My mother/father encourages me to diet to control my weight”
(two items). A scale score was created by taking the mean of these items.
Cronbach’s alpha was .75, and higher scores indicated greater perceived
parental involvement in dieting. However, because this scale was posi-
tively skewed, it was dichotomized at the 25th percentile for use in
multivariate analyses. Thus, participants with a scale score of 1 (corre-
sponding to the not at all response) were compared with those with higher
scores.
Peer environment. A single item, rated on a 4-point scale ranging from
1(not at all)to4(very much), assessed perceived friend dieting: “Many of
my friends diet to lose weight or keep from gaining weight.” There was
also a fifth option: “I don’t know.” For the regression analyses, responses
of somewhat and very much were combined because of relatively low
numbers endorsing very much, and in analyses conducted using this item,
respondents checking “I don’t know” were not included. Two items mea-
sured the frequency of being teased about weight and appearance on
5-point scales ranging from 1 (never)to5(at least once a week). These
were combined into a mean weight teasing score, with higher scores
indicating more frequent teasing.
Psychological variables. Self-esteem was assessed with a shortened
version of the Rosenberg Self-Esteem Inventory (Rosenberg, 1965), which
contained six items, such as “On the whole, I am quite satisfied with
myself,” and was rated on a 4-point scale ranging from 1 (strongly
disagree)to4(strongly agree). Higher scores indicated higher self-esteem,
and the scale had a Cronbach’s alpha of .78. Depressive mood was assessed
with a six-item scale assessing depressive symptoms (Kandel & Davies,
1982) such as “During the past 12 months, how often have you been
bothered or troubled by feeling unhappy, sad, or depressed?” Each item
was rated on a 3-point scale ranging from 1 (not at all)to3(very much),
and item responses were summed (Cronbach’s ␣⫽.75). Higher scores
indicated more depressive symptoms.
Data Analysis
Initially, Pearson product–moment correlations were used to identify
bivariate associations between Time 1 predictors and Time 2 body dissat-
isfaction, and relationships among predictors were examined for multicol-
linearity. To examine the predictive contribution of psychosocial factors
after Time 1 body dissatisfaction, we adopted a model-building approach
over several stages using multivariable linear regression analyses and
taking BMI and demographic variables into account. The first model
included Time 1 body dissatisfaction as the only independent variable.
BMI was added in the second model, and SES and ethnicity were added
subsequently to create a baseline model to which additional variables could
be added. Next, three separate models were created, which added (a) the
parent dieting environment score, (b) two peer environment variables, or
(c) two psychological variables to the baseline model. Finally, a reduced
model was created, in which Time 1 body satisfaction was used as a
predictor as well as all other variables, which had significance levels of p
.20 in the previous set of models. This criterion was set so as to ensure no
relevant variables were omitted.
All analyses were stratified by gender and age cohort, because of
previous research suggesting differences in several predictor variables
across gender or age groups. This approach allows the relationship between
all covariates and Time 2 body dissatisfaction to vary across both gender
and age, and it avoids the use of multiple, complex interaction terms in
each regression model. This strategy therefore identifies a unique set of
predictors and estimates their collective predictive power, within each
group.
Comparisons of Project EAT-I participants who did and did not respond
to Project EAT-II indicated several small but significant differences.
Project EAT-II participants were more likely than nonparticipants to be
female, in the older age group, European American, in upper SES catego-
ries, and in the healthy BMI range. Consequently, in all analyses, data were
weighted to adjust for differential response rates using the response pro-
891
BODY DISSATISFACTION IN ADOLESCENT GIRLS AND BOYS
pensity method (Little, 1986), in which the inverse of the estimated
probability that an individual responded at Time 2 was used as the weight.
Thus, all Time 2 estimates were generalizable to the population represented
by the original Project EAT-I sample. Analyses were conducted with SAS
8.2 (SAS Institute, 1999).
Results
Characteristics of Sample and Univariate Correlates of
Time 2 Body Dissatisfaction
Participant characteristics are summarized in Table 1. Univari-
ate correlations between putative Time 1 risk factor variables and
Time 2 body dissatisfaction were examined (see Table 2). All
proposed risk factors, except parent dieting environment and friend
dieting in the early adolescent boys and friend dieting in the older
girls, were correlated with Time 2 body satisfaction.
Prospective Predictors of Increase in Body Dissatisfaction
Early adolescent girls. Table 3 summarizes findings from the
multivariable linear regression analyses conducted for the early
adolescent girls in which Time 2 body dissatisfaction was the
dependent variable. Model 1, in which only Time 1 body dissat-
isfaction was entered, was significant. In Model 2, after Time 1
body dissatisfaction had been entered, BMI was a significant
unique predictor of Time 2 body dissatisfaction. In the baseline
model, in addition to Time 1 body dissatisfaction and BMI, SES,
being African American, and being of other–mixed ethnicity were
entered into the regression. However, none of the demographic
variables made unique contributions to the model. The first vari-
able added to the baseline model was parent dieting environment.
However, parent dieting did not contribute uniquely to the model.
Next, the two peer variables were added to the baseline model, and
friend dieting was observed to be a unique predictor of Time 2
body dissatisfaction. Finally, the two psychological variables were
added to the baseline model, and self-esteem was observed to be a
unique negative predictor of Time 2 body dissatisfaction.
In the reduced model shown at the bottom of Table 3, Time 1
body satisfaction, BMI, friend dieting, and self-esteem were en-
tered into the regression equation. It is interesting to note that Time
1 body dissatisfaction was not a significant unique predictor in this
model. However, in other respects, the results paralleled the pre-
vious analysis, with Time 1 BMI, friend dieting, and self-esteem
prospectively predicting Time 2 body satisfaction.
Early adolescent boys. Prospective predictors of Time 2 body
dissatisfaction in early adolescent boys were examined (see Table
4). Model 1, in which Time 1 body dissatisfaction only was
entered, was significant. When Time 1 body dissatisfaction and
BMI were entered together in Model 2, both variables contributed
uniquely to the variance in Time 2 body dissatisfaction. When the
baseline model was examined, in which Time 1 body dissatisfac-
tion, BMI, SES, and ethnicity were entered, Time 1 body dissat-
isfaction was no longer a unique predictor, but BMI and SES were
unique predictors, with higher BMI and lower SES predicting
greater Time 2 body dissatisfaction. When added to the baseline
model, parent dieting environment did not make a unique contri-
bution to Time 2 body dissatisfaction. However, when the peer
variables were entered, weight teasing was a unique, significant
predictor of Time 2 body dissatisfaction. Finally, when added to
the baseline model, neither self-esteem nor depression made a
unique contribution to the prediction of Time 2 body dissatisfac-
tion. In the reduced model of predictors of Time 2 body dissatis-
faction, Time 1 body dissatisfaction, BMI, SES, weight teasing,
and depression were entered into the analysis. BMI and SES were
Table 1
Means and Standard Deviations of Time 1 and Time 2 Body Dissatisfaction and Time 1 Body Mass Index, Parent Dieting
Environment, Friend Dieting, Weight Teasing, and Psychological Variables
Predictor variable
Scale
range
Early adolescents Middle adolescents
Girls Boys Girls Boys
MSDMSDMSDMSD
Body dissatisfaction Time 1 10–50 26.70 9.92 22.64 9.21 28.77 9.14 22.94 8.49
Body dissatisfaction Time 2 10–50 28.43 9.59 24.44 9.17 28.90 8.99 25.46 8.55
Body mass index 22.11 4.73 21.49 5.03 23.42 4.68 23.70 4.66
Parent dieting environment 1–4 1.80 0.84 1.88 0.85 1.67 0.68 1.74 0.70
Friend dieting 1–4 2.18 1.05 1.72 0.95 2.30 0.98 1.71 0.90
Weight teasing 1–5 2.13 1.20 1.97 1.15 1.94 1.06 1.91 1.02
Self-esteem 6–24 17.71 3.51 18.60 3.56 17.20 1.06 19.07 3.39
Depression 6–18 10.40 2.67 9.35 2.69 13.00 3.10 11.30 2.71
Table 2
Pearson’s Product–Moment Correlations Between Time 2 Body
Dissatisfaction and Putative Time 1 Risk Factor Variables
Time 1 variable
Early adolescents Middle adolescents
Girls Boys Girls Boys
Body dissatisfaction .35*** .22*** .48*** .42***
Body mass index .27*** .20*** .31*** .27***
Socioeconomic status .11* .22*** .15*** .12**
Parent dieting environment .18*** .10 .15*** .07*
Friend dieting .24*** .07 .05 .13***
Weight teasing .11* .16** .17*** .19***
Self-esteem .32*** .21*** .29*** .22***
Depression .24*** .11* .15*** .21***
* p .05. ** p .01. *** p .001.
892
PAXTON, EISENBERG, AND NEUMARK-SZTAINER
the only variables to make significant unique contributions to the
model.
Middle adolescent girls. In middle adolescent girls (see Table
5), Model 1, in which Time 1 body dissatisfaction was entered
alone, was significant. In Model 2, both body dissatisfaction and
BMI made significant unique contributions to Time 2 body dis-
satisfaction. In the baseline model, Time 1 body dissatisfaction,
BMI, SES (inversely), and being African American (inversely)
contributed uniquely to the variance. Neither parent dieting nor
Table 3
Summary of Multivariable Models for Early Adolescent Girls
Predictor variable SE t
Model 1: F(1, 420) 60.19***, R
2
.13
Body dissatisfaction 0.35 0.05 7.76***
Model 2: F(2, 410) 36.62***, R
2
.15
Body dissatisfaction 0.28 0.05 5.86***
Body mass index 0.40 0.10 4.21***
Baseline model: F(3, 387) 14.85***, R
2
.16
Body dissatisfaction 0.30 0.05 6.16***
Body mass index 0.32 0.10 3.18**
SES 0.10 0.36 0.29
African American 0.25 1.22 0.21
Other–mixed ethnicity 0.75 1.14 0.66
Baseline model plus parent environment: F(6, 383) 12.23***,
R
2
.16
Body dissatisfaction 0.30 0.05 5.94***
Body mass index 0.31 0.10 2.96**
SES 0.09 0.37 0.25
African American 0.28 1.22 0.23
Other–mixed ethnicity 0.87 1.15 0.76
Parent dieting environment 0.67 1.05 0.59
Baseline model plus peer environment: F(7, 288) 7.00***, R
2
.15
Body dissatisfaction 0.25 0.06 4.14***
Body mass index 0.19 0.12 1.64
SES 0.28 0.41 0.68
African American 0.87 1.37 0.63
Other–mixed ethnicity 0.24 1.27 0.19
Friend dieting 1.62 0.64 2.53*
Weight teasing 0.54 0.46 1.19
Baseline plus psychological variables: F(7, 372) 11.93***, R
2
.18
Body dissatisfaction 0.19 0.06 3.27***
Body mass index 0.30 0.10 2.97**
SES 0.35 0.37 0.95
African American 0.54 1.22 0.45
Other–mixed ethnicity 0.92 1.13 0.82
Self-esteem 0.52 0.17 3.11**
Depression 0.23 0.20 1.13
Reduced model: F(4, 305) 14.81***, R
2
.16
Body dissatisfaction 0.05 0.06 0.84
Body mass index 0.28 0.11 2.64**
Friend dieting 1.45 0.62 2.33*
Self-esteem 0.68 0.17 4.08***
Note. Degrees of freedom differ in each analysis because of differences
in missing data or, in the case of friend dieting, an “I don’t know” response.
SES socioeconomic status.
* p .05. ** p .01. *** p .001.
Table 4
Summary of Multivariable Models for Early Adolescent Boys
Predictor variable SE t
Model 1: F(1, 338) 18.17***, R
2
.05
Body dissatisfaction 0.23 0.05 4.26***
Model 2: F(2, 332) 9.34***, R
2
.05
Body dissatisfaction 0.15 0.06 2.43*
Body mass index 0.27 0.10 2.55*
Baseline model: F(5, 313) 7.50***, R
2
.11
Body dissatisfaction 0.06 0.06 1.02
Body mass index 0.39 0.10 3.92***
SES 1.10 0.45 2.44*
African American 0.32 1.36 0.23
Other–mixed ethnicity 1.02 1.17 0.88
Baseline model plus parent environment: F(6, 310) 6.38***,
R
2
.11
Body dissatisfaction 0.03 0.06 0.58
Body mass index 0.42 0.10 4.16***
SES 1.09 0.45 2.40*
African American 0.41 1.35 0.31
Other–mixed ethnicity 1.19 1.16 1.03
Parent dieting environment 0.15 1.17 0.13
Baseline model plus peer environment: F(7, 214) 6.32***, R
2
.17
Body dissatisfaction 0.05 0.06 0.85
Body mass index 0.28 0.13 2.17*
SES 1.04 0.47 2.24*
African American 2.07 1.67 1.24
Other–mixed ethnicity 1.49 1.24 1.20
Friend dieting 0.38 0.68 0.55
Weight teasing 1.48 0.50 2.98**
Baseline model plus psychological variables: F(7, 300) 5.64***,
R
2
.12
Body dissatisfaction 0.02 0.06 0.28
Body mass index 0.42 0.11 3.93***
SES 1.11 0.46 2.41*
African American 0.28 1.38 0.20
Other–mixed ethnicity 0.42 1.19 0.35
Self-esteem 0.09 0.17 0.55
Depression 0.30 0.21 1.43
Reduced model: F(5, 307) 7.86***, R
2
.11
Body dissatisfaction 0.00 0.06 0.07
Body mass index 0.39 0.11 3.56***
SES 1.22 0.42 2.90**
Weight teasing 0.49 0.46 1.05
Depression 0.29 0.19 1.53
Note. Degrees of freedom differ in each analysis because of differences
in missing data or, in the case of friend dieting, an “I don’t know” response.
SES socioeconomic status.
* p .05. ** p .01. *** p .001.
893
BODY DISSATISFACTION IN ADOLESCENT GIRLS AND BOYS
peer environment variables, when added to the baseline model,
made significant unique contributions to the prediction of Time 2
body dissatisfaction. However, when self-esteem and depression
were added to the baseline model, self-esteem made a significant
unique contribution to the prediction of Time 2 body dissatisfac-
tion. In the reduced model for middle adolescent girls, Time 1
body dissatisfaction, BMI, SES, being African American, being of
other–mixed ethnicity, friend dieting, weight teasing, and self-
esteem were entered into the model. In this analysis, significant
unique predictors of Time 2 body dissatisfaction were Time 1 body
dissatisfaction, BMI, and SES.
Middle adolescent boys. Finally, predictors of Time 2 body
dissatisfaction in middle adolescent boys were examined (see
Table 6). Model 1 was significant and in Model 2, both Time 1
body dissatisfaction and BMI were significant unique predictors of
Time 2 body dissatisfaction. When the baseline model was exam-
ined, Time 1 body dissatisfaction and BMI positively contributed
to the model, whereas being African American and being of
other–mixed ethnicity were unique inverse predictors of Time 2
body dissatisfaction. When added to the baseline model, neither
parent dieting nor peer variables made significant contributions to
the variance explained. However, when the psychological vari-
ables were entered, depression made a unique contribution to the
model. The variables entered into the reduced model for middle
adolescent boys were Time 1 body dissatisfaction, BMI, being
African American, being of other–mixed ethnicity, and depression.
Each of these variables made significant unique contributions to
the prediction of Time 2 body dissatisfaction.
Discussion
This study examined prospective risk factors for increases in
body dissatisfaction in girls and boys over a relatively long time
period of 5 years, in two different developmental phases, from
early to middle adolescence and from middle to late adolescence.
Predictors of Time 2 body dissatisfaction were examined in each
group separately in order to facilitate identification of different
patterns of predictors. Time 1 body dissatisfaction, when entered
alone, was a significant prospective predictor of Time 2 body
dissatisfaction in all groups. However, in the reduced models,
Time 1 body dissatisfaction was not a significant predictor in early
adolescent girls and boys. As hypothesized and consistent with
previous research, BMI was a predictor of increases in body
dissatisfaction in all samples. Of the demographic variables exam-
ined, SES inversely predicted Time 2 body dissatisfaction in early
adolescent boys and middle adolescent girls. Partially consistent
with our hypothesis, being African American predicted a lower
increase in body dissatisfaction in both boys and girls in the middle
but not in the early adolescent cohort. Parent dieting was not a
prospective predictor in any group. However, in partial support of
our hypothesis, in early adolescence, friend dieting and weight
teasing predicted an increase in body dissatisfaction in girls and
boys, respectively, but this was not the case in the older cohort.
Partially consistent with our hypothesis, depression positively pre-
dicted Time 2 body dissatisfaction in middle adolescent boys.
Finally, self-esteem inversely predicted increases in body dissat-
isfaction in both young and middle adolescent girls.
In this multivariable analysis of prospective risk factors of Time
2 body dissatisfaction, Time 1 body dissatisfaction was controlled
Table 5
Summary of Multivariable Models for Middle Adolescent Girls
Predictor variable SE t
Model 1: F(1, 911) 278.46***, R
2
.23
Body dissatisfaction 0.46 0.03 16.69***
Model 2: F(2, 897) 155.83***, R
2
.26
Body dissatisfaction 0.40 0.02 13.49***
Body mass index 0.32 0.06 5.77***
Baseline model: F(5, 882) 69.28***, R
2
.28
Body dissatisfaction 0.37 0.03 12.28***
Body mass index 0.32 0.06 5.78***
SES 0.85 0.22 3.83***
African American 2.04 0.79 2.58**
Other–mixed ethnicity 0.10 0.64 1.56
Baseline model plus parent environment: F(6, 876) 57.95***,
R
2
.28
Body dissatisfaction 0.37 0.03 12.37***
Body mass index 0.31 0.06 5.45***
SES 0.88 0.22 3.96***
African American 1.99 0.80 2.47**
Other–mixed ethnicity 0.96 0.64 1.50
Parent environment 0.34 0.59 0.58
Baseline model plus peer environment: F(7, 790) 45.59***, R
2
.29
Body dissatisfaction 0.41 0.03 12.32***
Body mass index 0.25 0.06 4.13***
SES 0.90 0.23 3.85***
African American 1.58 0.89 1.77
Other–mixed ethnicity 0.78 0.67 1.16
Friend dieting 0.49 0.35 1.43
Weight teasing 0.37 0.26 1.38
Baseline model plus psychological variables: F(7, 864) 46.37***,
R
2
.27
Body dissatisfaction 0.32 0.03 9.29***
Body mass index 0.32 0.06 5.57***
SES 0.85 0.22 3.82***
African American 1.97 0.80 2.48*
Other–mixed ethnicity 0.79 0.65 1.23
Self-esteem 0.22 0.10 2.22*
Depression 0.04 0.11 0.40
Reduced model: F(8, 779) 39.24***, R
2
.29
Body dissatisfaction 0.39 0.04 10.33***
Body mass index 0.25 0.06 4.21***
SES 0.92 0.24 3.91***
African American 1.60 0.90 1.79
Other–mixed ethnicity 0.67 0.68 0.98
Friend dieting 0.50 0.35 1.43
Weight teasing 0.26 0.28 0.92
Self-esteem 0.10 0.10 1.04
Note. Degrees of freedom differ in each analysis because of differences
in missing data or, in the case of friend dieting, an “I don’t know” response.
SES socioeconomic status.
* p .05. ** p .01. *** p .001.
894
PAXTON, EISENBERG, AND NEUMARK-SZTAINER
for by entering it into each of the models examined. Notably,
however, in the reduced models examined in the early adolescent
cohorts, Time 1 body dissatisfaction was not a unique predictor of
Time 2 body dissatisfaction. This finding suggests that at this stage
of development when girls and boys are typically still growing
physically and developing their identity, body dissatisfaction is not
yet stable, and the development of body dissatisfaction is influ-
enced to a greater extent by other physical, social, and individual
attributes partially associated with body dissatisfaction at Time 1.
This is not the case, however, for the middle adolescent cohort, in
which Time 1 body dissatisfaction makes a strong contribution to
the variance in Time 2 body dissatisfaction in the multivariable
analyses. These findings support prevention efforts earlier rather
than later in adolescence.
Consistent with most previous research (Field et al., 2001;
Halpern et al., 1999; Ohring et al., 2002), BMI was a strong and
consistent predictor of increases in body dissatisfaction in both
developmental phases examined. The thin beauty ideal is propa-
gated through media that reach all age groups and are consistent
and pervasive (Levine & Harrison, 2004), and thinness is associ-
ated with popularity and sexual, relationship, and material suc-
cess—issues pertinent to boys and girls across both developmental
periods we examined. On the other hand, research has shown that
overweight people suffer a variety of poor social outcomes, in-
cluding job discrimination, social exclusion, mistreatment by doc-
tors, public ridicule, and rejection of college admission and fund-
ing (Crandall, 1994; Latner, Stunkard, & Wilson, 2005; Puhl &
Brownell, 2001; Schwartz, Chambliss, Brownell, Blair, & Billing-
ton, 2003). Against this cultural backdrop, it is not surprising that
departure from this ideal is associated with increases in body
dissatisfaction in our samples.
SES and ethnicity, most likely proxy variables for subcultural
norms and attitudes, prospectively predicted increases in body
dissatisfaction in different ways in the two cohorts. Lower SES
prospectively predicted Time 2 body dissatisfaction in early ado-
lescent boys and middle adolescent girls. Lower self-esteem is
associated with lower SES (e.g., Rhodes, Roffman, Reddy,
Fredriksen, & Way, 2004; Twenge & Campbell, 2002), which may
also affect an adolescent’s view of his or her body. Particularly in
relation to the girls, a further explanation is that girls in lower SES
may be less able to afford fashionable clothing that may impinge
on body image more widely. It is not clear, however, why SES
plays a role at different developmental phases in girls and boys,
and further research is required to clarify the nature of these
relationships.
African American ethnicity emerged as a protective factor
against increases in body dissatisfaction in middle adolescence.
These findings are consistent with cross-sectional studies that have
observed that body satisfaction is higher in African American girls
and boys in comparison with other ethnic groups in the U.S. (Akan
& Grillo, 1995; Kelly et al., 2005; Neumark-Sztainer, Croll, et al.,
2002; Neumark-Sztainer et al., 1999). The larger body size ideal in
the African American subculture is more consistent with the nat-
ural shape of young women from middle to late adolescence, and
consequently the extent of departure from this body shape ideal is
likely to be less, offering protection against increases in body
dissatisfaction. Less is known about body image among boys and
men, particularly among males from different racial– ethnic back-
grounds. Research has suggested that African American males are
less likely than European American males to consider themselves
overweight (Rand & Kuldau, 1990) and that, overall, males tend to
desire more muscular shapes than they describe themselves as
having (Pope et al., 2000). African American males may be more
Table 6
Summary of Multivariable Models for Middle Adolescent Boys
Predictor variable SE t
Model 1: F(1, 733) 153.72***, R
2
.17
Body dissatisfaction 0.41 0.03 12.40***
Model 2: F(2, 727) 96.69***, R
2
.21
Body dissatisfaction 0.35 0.03 10.21***
Body mass index 0.36 0.06 5.83***
Baseline model: F(5, 714) 45.72***, R
2
.24
Body dissatisfaction 0.32 0.03 9.50***
Body mass index 0.38 0.06 6.10***
SES 0.24 0.25 0.97
African American 3.01 0.85 3.54***
Other–mixed ethnicity 2.00 0.76 2.65**
Baseline model plus parent environment: F(6, 710) 39.21***,
R
2
.25
Body dissatisfaction 0.34 0.03 9.74***
Body mass index 0.38 0.06 6.11***
SES 0.20 0.25 0.83
African American 2.94 0.84 3.48***
Other–mixed ethnicity 2.04 0.75 2.73**
Parent dieting environment 0.28 0.65 0.43
Baseline model plus peer environment: F(7, 606) 26.66***, R
2
.24
Body dissatisfaction 0.35 0.04 8.72***
Body mass index 0.28 0.07 4.05***
SES 0.16 0.27 0.58
African American 0.86 1.00 0.86
Other–mixed ethnicity 2.48 0.82 3.02**
Friend dieting 0.49 0.40 1.23
Weight teasing 0.36 0.31 1.16
Baseline model plus psychological variables: F(7, 703) 35.14***,
R
2
.26
Body dissatisfaction 0.30 0.04 7.69***
Body mass index 0.36 0.06 5.82***
SES 0.30 0.25 1.21
African American 2.95 0.88 3.33***
Other–mixed ethnicity 2.09 0.75 2.79**
Self-esteem 0.06 0.10 0.62
Depression 0.29 0.12 2.46*
Reduced model: F(5, 716) 49.63***, R
2
.26
Body dissatisfaction 0.31 0.04 8.58***
Body mass index 0.37 0.06 5.99***
African American 2.88 0.83 3.46***
Other–mixed ethnicity 2.42 0.65 3.73***
Depression 0.33 0.11 3.03**
Note. Degrees of freedom differ in each analysis because of differences
in missing data or, in the case of friend dieting, an “I don’t know” response.
SES socioeconomic status.
* p .05. ** p .01. *** p .001.
895
BODY DISSATISFACTION IN ADOLESCENT GIRLS AND BOYS
likely to fit their ideal as they move into young adulthood, whereas
those of other–mixed ethnicity, mainly Asian and Hispanic in this
research, may be less likely to do so.
The parent dieting environment measure did not predict in-
creases in body dissatisfaction in either cohort in this research. In
a study by Field and colleagues (2001) that observed more fre-
quent development of weight concerns when thinness was impor-
tant to parents, the follow-up period was relatively short (1 year).
Parent dieting environment factors may be more proximal influ-
ences and less relevant over a longer period. In addition, in
cross-sectional studies, parent encouragement to diet (a component
of parent dieting environment) has been closely associated with
BMI (Wertheim et al., 1999, 2002). Thus, it may be the wider
consequences of a higher BMI that are particularly important in the
increase in body dissatisfaction over 5 years, rather than parent
dieting attitudes and behaviors in particular. Further, from a de-
velopmental perspective, it would be anticipated that parent influ-
ences would be especially pertinent during childhood and that by
adolescence, factors external to the family may become more
salient in their effect.
Peer environment factors were prospective predictors of in-
creases in body dissatisfaction from early to middle adolescence.
In the early adolescent cohort of girls, friend dieting predicted
increases in body dissatisfaction, a finding consistent with cross-
sectional (e.g., Jones et al., 2004; Paxton et al., 1999) and longi-
tudinal (Jones, 2004) research in which friend dieting concerns and
appearance conversations have been associated with body dissat-
isfaction. Mechanisms involved in the relationship between
appearance-concerned friendship environment and body dissatis-
faction have yet to be fully elucidated. However, perceived friend
concern with weight and conversations about appearance have
been observed to be associated with enhanced internalization of
the thin ideal and body comparisons (Jones, 2004; Jones et al.,
2004; Schutz, Paxton, & Wertheim, 2002). Thus, the effect of
friend dieting on increases in body dissatisfaction from early to
middle adolescence may be mediated by these attributes.
It is notable that friend dieting was not a significant predictor of
increases in body dissatisfaction in the middle adolescent cohort.
Early to middle adolescence is a developmental phase especially
associated with the need to feel accepted by peers, which may
intensify the influence of peer factors on body image during this
time (Brown, 2004; Perry, 2001). In addition, for many of the
middle adolescent cohort, participants’ friendship environment
may have changed substantially over the 5-year period as a con-
sequence of leaving school and moving into the young adult world.
Changes in peer environments over middle to later adolescence
may contribute to the weakening of peer influences on increases in
body dissatisfaction during this developmental phase.
Weight teasing was a unique predictor of increases in body
dissatisfaction in early adolescent boys when entered with the
baseline model. However, its effect diminished when entered with
depression in the reduced model, suggesting a mediating relation-
ship between the two. Weight teasing not only asserts the norms of
attractiveness in a group but it also asserts disapproval, criticism,
and rejection. In light of this reality, dissatisfaction with the
apparent source of the problem, the body, would seem a natural
consequence, especially during this developmental phase during
which peer acceptance and affiliation is a central developmental
task. Observation of a prospective role for weight teasing earlier
but not later in adolescence is consistent with previous research
finding weight teasing prospectively predicted bulimic symptoms
in 7th-grade girls but not in older adolescents (Wertheim et al.,
2001). However, weight teasing was also strongly correlated with
Time 2 body dissatisfaction in the middle adolescent cohorts, and
it is likely it was not a unique predictor because of its close
association with BMI (Neumark-Sztainer, Falkner, et al., 2002).
Further investigation is required to identify these relationships.
The final variables examined were the psychological variable,
self-esteem, and depression. Although low self-esteem is typically
correlated with body dissatisfaction, it has not previously been
identified as a prospective risk factor for body dissatisfaction in
girls. Identification of low self-esteem as a predictor in this study
may relate to the longer follow-up period. The mechanisms
through which self-esteem may operate need further exploration.
However, low self-esteem may contribute to an increase in nega-
tive self-evaluation generally and in negative evaluation of the
body particularly, over this time frame. Low self-esteem may also
be a characteristic that, over an extended period, makes a girl
vulnerable to more proximal pressures, such as environmental
pressures to be thin, that have been shown to predict increases in
body dissatisfaction over a 1-year period (Stice & Whitenton,
2002). Further, low self-esteem may contribute to a girl being
more likely to internalize cultural ideals of success, including the
thin beauty ideal, and to engage in body comparisons, attitudes,
and processes that increase vulnerability to increases in body
dissatisfaction (Durkin & Paxton, 2002).
Although Time 1 depression was correlated with Time 2 body
dissatisfaction in all groups, it was only a unique predictor in
middle adolescent boys, consistent with the research of Presnell
and colleagues (2004). Thus, negative mood rather than a more
general sense of self-worth appears particularly relevant to change
in body dissatisfaction in this group. Depressive symptoms typi-
cally involve negative feelings about the self that are likely, over
time, to include negative feelings toward body image. As with
self-esteem, depression may increase vulnerability to real or per-
ceived criticism and pressures to achieve the ideal body.
The amount of variance explained by the final reduced model
varied widely between the early and middle adolescent cohorts,
and both Time 1 body dissatisfaction and additional predictor
variables contributed less to the prediction of Time 2 body dissat-
isfaction in early adolescents. Thus, although this research indi-
cates that the variables described above played a part in increases
in body dissatisfaction, in the early adolescent cohort especially,
other factors that we did not assess were also clearly relevant.
Other social, biological, and psychological factors, such as media
exposure and pressure to be thin, internalization of the sociocul-
tural thin beauty ideal, body comparison, and perfectionism, may
also be prospective predictors. Further research is required to
explore these factors.
A principal applied reason for risk factor research is to identify
issues of relevance to prevention interventions. The present find-
ings indicate the importance of intervention at least prior to middle
adolescence if not earlier, before body dissatisfaction has become
a stable feature of a person’s make-up. The crucial role of BMI in
the Western social context as a prospective risk factor for increases
in body dissatisfaction across teenage years highlights the impor-
tance of healthy weight control for both physical and mental
health. Unfortunately, to date, only limited success has been re-
896
PAXTON, EISENBERG, AND NEUMARK-SZTAINER
ported in public health interventions for healthy weight mainte-
nance (McGuire, Wing, Klem, & Hill, 1999). The fact that higher
SES and being African American appear to be protective factors is
not as negative for prevention as it may first appear. These demo-
graphic characteristics are likely to be markers for attitudinal
norms, and if acceptance of a greater diversity in body ideals can
be endorsed in one subculture, they can potentially be endorsed by
another. In the early adolescent cohort, the present research sug-
gests the importance of providing supportive peer environments.
Few interventions to date appear to have made this a focus.
Finally, interventions that raise self-esteem and reduce depression
are likely to be valuable in reducing body dissatisfaction. Consis-
tent with this proposal are the findings of O’Dea and Abraham
(2000), who observed positive effects on body image in a program
that emphasized building self-esteem. Early identification and
treatment of depression may be especially beneficial in older boys.
The current study has a number of strengths that enhance our
ability to draw meaningful conclusions from our findings. The
study population was larger and more diverse than other prospec-
tive studies examining predictors of body dissatisfaction (Wer-
theim et al., 2004), thus allowing for extrapolations to both genders
and to lower income and ethnically diverse populations. The broad
array of measures facilitated testing of a multivariable, psychoso-
cial model of increases in body dissatisfaction. The 5-year
follow-up period enabled an examination of longer term risk
factors for increases in body dissatisfaction. Finally, the two-
cohort design facilitated an examination of risk factors at two key
developmental transitions, early to middle adolescence and middle
to late adolescence.
However, the study also has limitations that need to be taken
into account in interpreting the findings. There was substantial
attrition between the two testing occasions, which may introduce
bias into the data. Specifically, at Time 2, the sample was more
likely to be European American and in the upper SES categories
compared with the Time 1 sample. This likely is due to the greater
mobility of participants in the lower SES and non-European Amer-
ican groups, who were heavily represented in the original Project
EAT sample. Population weights reflecting the original sample
were therefore used in all analyses to address this shortcoming and
allow inferences to the original study population. Nevertheless,
lower rates of attrition would have been preferable given the
limitations of statistical adjustments and to allow for greater power
from a larger sample size. Furthermore, although the original study
population was large and diverse, it was not representative of all
U.S. adolescents; thus, extrapolations to populations outside of the
sample should be made cautiously. A further limitation relates to
the range of variables assessed and the instruments used. Although
a range of potential predictor variables selected on theoretical and
empirical grounds was used in this research, as indicated above,
important predictors may have been omitted, such as media pres-
sure and internalization of the thin ideal. Further, being part of a
larger epidemiological project, the measures were kept as short as
possible. In particular, in future research, examination of peer
influences using a more extensive instrument would be beneficial.
Except in the case of BMI, which was measured, only self-report
assessments were used. Ideally, more objective measures would be
obtained, but compromises are required to obtain a substantial
sample. Finally, this research aimed to identify prospective pre-
dictors of body dissatisfaction. It is likely that these predictors
interact in a range of complex ways, some variables potentially
mediating or moderating the impact of other variables on body
dissatisfaction. For example, effects may differ according to eth-
nicity or age. Future research, using different analytical strategies,
is required to explore any mediating pathways or interaction ef-
fects among the predictors identified here.
In conclusion, the current research identified prospective pre-
dictors of increases in body dissatisfaction over a 5-year period in
adolescent girls and boys. During the early adolescent phase, there
is less stability in body dissatisfaction and greater influence of peer
factors in its development. Over this 5-year time frame, self-
esteem is a prospective risk factor in girls but not in boys, for
whom depression was more influential in the middle adolescent
cohort. Finally, the present research highlights the need to consider
demographic factors in longitudinal research such as being African
American and being from a higher SES as protective at different
stages of development.
References
Akan, G. E., & Grillo, C. M. (1995). Socio-cultural influences on eating
attitudes and behaviors, body image, and psychological functioning: A
comparison of African America, Asian American, and Caucasian college
women. International Journal of Eating Disorders, 18, 181–187.
Barr Taylor, C., Sharpe, T., Shisslak, C., Bryson, S., Estes, L. S., Gray, N.,
et al. (1998). Factors associated with weight concerns in adolescent girls.
International Journal of Eating Disorders, 24, 31– 42.
Berndt, T. J., & Hestenes, S. L. (1996). The developmental course of social
support: Family and peers. In L. Smolak, M. P. Levine, & R. Striegel-
Moore (Eds.), The developmental psychopathology of eating disorders
(pp. 77–106). Mahwah, NJ: Erlbaum.
Birch, L. L., Fisher, J. O., & Davison, K. K. (2003). Learning to overeat:
Maternal use of restrictive feeding practices promotes girls’ eating in the
absence of hunger. American Journal of Clinical Nutrition, 78, 215–220.
Brown, B. B. (2004). Adolescents’ relationships with peers. In R. M.
Lerner & L. Steinberg (Eds.), Handbook of adolescent psychology (2nd
ed., pp. 363–394). Hoboken, NJ: Wiley.
Byely, L., Archibald, A. B., Graber, J., & Brooks-Gunn, J. (2000). A
prospective study of familial and social influences on girls’ body image
and dieting. International Journal of Eating Disorders, 28, 155–164.
Cafri, G., Thompson, J. K., Ricciardelli, L., McCabe, M., Smolak, L., &
Yesalis, C. E. (2005). Pursuit of the muscular ideal: Physical and
psychological consequences and putative risk factors. Clinical Psychol-
ogy Review, 25, 215–239.
Cash, T. F. (2002). Cognitive behavioral perspectives on body image. In
T. F. Cash & T. Pruzinski (Eds.), Body image: A handbook of theory,
research, and clinical practice (pp. 38 46). New York: Guilford Press.
Cattarin, J. A., & Thompson, J. K. (1994). A three-year longitudinal study
of body image, eating concerns, and general psychological functioning
in adolescent females. Eating Disorders: Journal of Treatment and
Prevention, 2, 114 –125.
Cohane, G. H., & Pope, H. G., Jr. (2001). Body image in boys: A review
of the literature. International Journal of Eating Disorders, 29, 373–379.
Crandall, C. S. (1994). Prejudice against fat people: Ideology and self-
interest. Journal of Personality and Social Psychology, 66, 882– 894.
Durkin, S. J., & Paxton, S. J. (2002). Predictors of vulnerability to reduced
body image satisfaction and psychological well-being in response to
exposure to idealized female media images in adolescent girls. Journal
of Psychosomatic Research, 53, 995–1005.
Eisenberg, M. E., Neumark-Sztainer, D., & Story, M. (2003). Associations
of weight-based teasing and emotional well-being among adolescents.
Archives of Pediatric and Adolescent Medicine, 157, 733–738.
Field, A. F., Carmago, C. A., Taylor, C. B., Berkey, C. S., Roberts, S. B.,
897
BODY DISSATISFACTION IN ADOLESCENT GIRLS AND BOYS
& Colditz, G. A. (2001). Peer, parent, and media influences on the
development of weight concerns and frequent dieting among preadoles-
cent and adolescent girls. Pediatrics, 107, 54 60.
Fisher, J. O., & Birch, L. L. (1999). Restricting access to palatable foods
affects children’s behavioral response, food selection, and intake. Amer-
ican Journal of Clinical Nutrition, 69, 1264 –1272.
Fisher, J. O., & Birch, L. L. (2002). Eating in the absence of hunger and
overweight in girls from 5 to 7 years of age. American Journal of
Clinical Nutrition, 76, 226 –231.
Fulkerson, J. A., McGuire, M. T., Neumark-Sztainer, D., Story, M.,
French, S. A., & Perry, C. L. (2002). Weight-related attitudes and
behaviors of adolescent boys and girls who are encouraged to diet by
their mothers. International Journal of Obesity, 26, 1579 –1587.
Halpern, C. T., Udry, J. R., Campbell, B., & Suchindran, C. (1999). Effects
of body fat on weight concerns, dating, and sexual activity: A longitu-
dinal analysis of Black and White adolescent girls. Developmental
Psychology, 35, 721–736.
Holsen, I., Kraft, P., & Roysamb, E. (2001). The relationship between body
image and depressed mood in adolescence: A 5-year longitudinal panel
study. Journal of Health Psychology, 6, 613– 627.
Jones, D. C. (2004). Body image among adolescent girls and boys: A
longitudinal study. Developmental Psychology, 40, 823– 835.
Jones, D. C., Vigfusdottir, T. H., & Lee, Y. (2004). Body image and the
appearance culture among adolescent girls and boys: An examination of
friend conversations, peer criticism, appearance magazines, and inter-
nalization of appearance ideals. Journal of Adolescent Research, 19,
323–339.
Kandel, D. B., & Davies, M. (1982). Epidemiology of depressive mood in
adolescents: An empirical study. Archives of General Psychiatry, 39,
1205–1212.
Keery, H., Eisenberg, M. E., Boutelle, K., Neumark-Sztainer, D., & Story,
M. (in press). Relationships between maternal and adolescent weight-
related behaviors and concerns: The role of perception. Journal of
Psychosomatic Research.
Kelly, A. M., Wall, M., Eisenberg, M. E., Story, M., & Neumark-Sztainer,
D. (2005). Adolescent girls with high body satisfaction: Who are they
and what can they teach us? Journal of Adolescent Health, 37, 391–396.
Latner, J. D., Stunkard, A. J., & Wilson, G. T. (2005). Stigmatized
students: Age, sex, and ethnicity effects in the stigmatization of obesity.
Obesity Research, 13, 1226 –1231.
Levesque, R. P. (1993). The romantic experience of adolescents in satis-
fying love relationships. Journal of Youth and Adolescence, 22, 219
252.
Levine, M. P., & Harrison, K. (2004). Media’s role in the perpetuation and
prevention of negative body image and disordered eating. In J. K.
Thompson (Ed.), Handbook of eating disorders and obesity (pp. 695–
717). Hoboken, NJ: Wiley.
Little, R. J. A. (1986). Survey nonresponse adjustments for estimates of
means. International Statistical Review, 54, 137–139.
Martin, G. C., Wertheim, E. H., Prior, M., Smart, D., Sanson, A., &
Oberklaid, F. (2000). A longitudinal study of the role of childhood
temperament in the later development of eating concerns. International
Journal of Eating Disorders, 27, 150 –162.
McCabe, M. P., & Ricciardelli, L. A. (2003). Body image and strategies to
lose weight and increase muscle among boys and girls. Health Psychol-
ogy, 22, 39 46.
McGuire, M. T., Wing, R. R., Klem, M. L., & Hill, J. O. (1999). Behavioral
strategies of individuals who have maintained long-term weight losses.
Obesity Research, 7, 334 –341.
Neumark-Sztainer, D., Croll, J., Story, M., Hannan, P. J., French, S., &
Perry, C. (2002). Ethnic/racial differences in weight-related concerns
and behaviors among adolescent girls and boys: Findings from Project
EAT. Journal of Psychosomatic Research, 53, 963–974.
Neumark-Sztainer, D., Falkner, N., Story, M., Perry, C., Hannan, P., &
Mulert, S. (2002). Weight-teasing among adolescents: Correlations with
weight status and disordered eating behaviors. International Journal of
Obesity, 1, 123–131.
Neumark-Sztainer, D., Story, M., Faibisch, L., Ohlson, J., & Adamiak, M.
(1999). Issues of self-image among overweight African American and
Caucasian adolescent girls: A qualitative study. Journal of Nutrition
Education, 31, 311–320.
Neumark-Sztainer, D., Story, M., Hannan, P. J., & Croll, J. (2002). Over-
weight status and eating patterns among adolescents: Where do youth
stand in comparison to the Healthy People 2010 Objectives. American
Journal of Public Health, 92, 844 851.
Neumark-Sztainer, D., Story, M., Hannan, P. J., Perry, C., & Irving, L. M.
(2002). Weight-related concerns and behaviors among overweight and
non-overweight adolescents: Implications for preventing weight-related
disorders. Archives of Pediatrics and Adolescent Medicine, 156, 171–
178.
Neumark-Sztainer, D., Wall, M., Story, M., & Perry, C. (2003). Correlates
of unhealthy weight control behaviors among adolescent girls and boys:
Implications for the primary prevention of disordered eating. Health
Psychology, 22, 88 –98.
Nichter, M. (2000). Fat talk: What girls and their parents say about
dieting. Cambridge, MA: Harvard University Press.
O’Dea, J., & Abraham, S. (2000). Improving the body image, eating
attitudes, and behaviors of young male and female adolescents: A new
educational approach that focuses on self-esteem. International Journal
of Eating Disorders, 28, 43–57.
Ohring, R., Graber, J. A., & Brooks-Gunn, J. (2002). Girls’ recurrent and
concurrent body dissatisfaction: Correlates and consequences over 8
years. International Journal of Eating Disorders, 31, 404 415.
Paxton, S. J., Norris, M., Wertheim, E. H., Durkin, S. J., & Anderson, J.
(2005). Body dissatisfaction, dating, and importance of thinness to
attractiveness in adolescent girls and boys. Sex Roles: A Journal of
Research, 53, 663– 675.
Paxton, S. J., Schutz, H. K., Wertheim, E. H., & Muir, S. L. (1999).
Friendship clique and peer influences on body image attitudes, dietary
restraint, extreme weight-loss behaviors, and binge eating in adolescent
girls. Journal of Abnormal Psychology, 108, 255–266.
Perry, C. L. (2001). Preadolescent and adolescent influences on health. In
B. D. Smedley & S. L. Syme (Eds.), Promoting health: Intervention
strategies from social and behavioral research (pp. 217–253). Wash-
ington, DC: National Academy Press, Institute of Medicine, Division of
Health Promotion and Disease Prevention.
Pingitore, R., Spring, B., & Garfield, D. (1997). Gender differences in body
satisfaction. Obesity Research, 5, 402– 409.
Pope, H. G., Gruber, A. J., Mangweth, B., Bureau, B., deCol, C., Juvent,
R., & Hudsen, J. I. (2000). Body image perception among men in three
countries. American Journal of Psychiatry, 157, 1297–1301.
Presnell, K., Bearman, S. K., & Stice, E. (2004). Risk factors for body
dissatisfaction in adolescent boys and girls: A prospective study. Inter-
national Journal of Eating Disorders, 36, 389 401.
Puhl, R., & Brownell, K. D. (2001). Bias, discrimination, and obesity.
Obesity Research, 9, 788 805.
Rand, C. S. W., & Kuldau, J. M. (1990). The epidemiology of obesity and
self-defined weight problem in the general population: Gender, race, age
and social class. International Journal of Eating Disorders, 9, 329 –343.
Rhodes, J., Roffman, J., Reddy, R., Fredriksen, K., & Way, N. (2004).
Changes in self-esteem during the middle school years: A latent growth
curve analysis study of individual and contextual factors. Journal of
School Psychology, 42, 243–261.
Ricciardelli, L. A., & McCabe, M. P. (2001). Dietary restraint and negative
affect as mediators of body dissatisfaction and bulimic behavior in
adolescent girls and boys. Behaviour Research and Therapy, 39, 1317–
1328.
Ricciardelli, L. A., & McCabe, M. P. (2003). A longitudinal analysis of the
898
PAXTON, EISENBERG, AND NEUMARK-SZTAINER
role of biopsychosocial factors predicting body change strategies among
adolescent boys. Sex Roles: A Journal of Research, 48, 349 –359.
Robinson, T. N., Kiernan, M., Matheson, D. M., & Haydel, K. F. (2001).
Is parental control over children’s eating associated with childhood
obesity? Results from a population-based sample of third graders. Obe-
sity Research, 9, 306 –312.
Rosenberg, M. (1965). Society and adolescent self-image. Princeton, NJ:
Princeton University.
SAS Institute. (1999). Proprietary Software Release 8.2 (TS2M0) [Com-
puter software]. Cary, NC: Author.
Schutz, H. K., Paxton, S. J., & Wertheim, E. H. (2002). Investigation of
body comparison among adolescent girls. Journal of Applied Social
Psychology, 32, 1906 –1937.
Schwartz, M. B. Chambliss, H. O., Brownell, K. B., Blair, S. N., &
Billington, C. (2003). Weight bias among health professionals special-
izing in obesity. Obesity Research, 11, 1033–1039.
Spangler, D. L. (2002). Testing the cognitive model of eating disorders:
The role of dysfunctional beliefs about appearance. Behavior Therapy,
33, 87–105.
Spruijt-Metz, D., Lindquist, C. H., Birch, L. L., Fisher, J. O., & Goran,
M. I. (2002). Relation between mothers’ child-feeding practices and
children’s adiposity. American Journal of Clinical Nutrition, 75, 581–
586.
Stice, E. (1994). Review of the evidence for a socio-cultural model of
bulimia nervosa and an exploration of the mechanisms of action. Clin-
ical Psychology Review, 14, 633– 661.
Stice, E. (2001). A prospective test of the dual-pathway model of bulimic
pathology: Mediating effects of dieting and negative affect. Journal of
Abnormal Psychology, 110, 124 –135.
Stice, E. (2002). Risk and maintenance factors for eating pathology: A
meta-analytic review. Psychological Bulletin, 128, 825– 848.
Stice, E., & Bearman, S. K. (2001). Body image and eating disturbances
prospectively predict increases in depressive symptoms in adolescent
girls: A growth curve analysis. Developmental Psychology, 37, 597–
607.
Stice, E., & Whitenton, K. (2002). Risk factors for body dissatisfaction in
adolescent girls: A longitudinal investigation. Developmental Psychol-
ogy, 38, 669 678.
Striegel-Moore, R. H., Schreiber, G. M., Lo, A., Crawford, P., Pbarzanek,
E., & Rodin, J. (2000). Eating disorder symptoms in a cohort of 11 to
16-year-old Black and White girls: The NHLBI Growth And Health
Study. International Journal of Eating Disorders, 27, 49 66.
Tanner, J. M. (1989). Fetus into man: Physical growth from conception to
maturity (2nd ed). Cambridge, MA: Harvard University Press.
Thompson, J. K., Heinberg, L. J., Altabe, M., & Tantleff-Dunn, S. (1999).
Exacting beauty: Theory, assessment, and treatment of body image
disturbance. Washington, DC: American Psychological Association.
Twenge, J. M., & Campbell, W. K. (2002). Self-esteem and socio-
economic status: A meta-analytic review. Personality and Social Psy-
chology Review, 6, 59 –71.
van den Berg, P., Thompson, J. K., Obremski-Brandon, K., & Coovert, M.
(2002). The Tripartite Influence model of body image and eating dis-
turbance: A covariance structure modeling investigation testing the
mediational role of appearance comparison. Journal of Psychosomatic
Research, 53, 1007–1020.
Wade, T., & Lowes, J. (2002). Variables associated with disturbed eating
habits and the overvaluation of ideas about the personal implications of
body shape and weight in a female adolescent population. International
Journal of Eating Disorders, 32, 39 45.
Wertheim, E. H., Koerner, J., & Paxton, S. J. (2001). Longitudinal predic-
tors of restrictive eating and bulimic behavior in adolescent girls. Jour-
nal of Youth and Adolescence, 30, 69 81.
Wertheim, E. H., Martin, G., Prior, M., Sanson, A., & Smart, D. (2002).
Parent influences in the transmission of eating and weight related values
and behaviors. Eating Disorders: The Journal of Treatment and Pre-
vention, 10, 329 –342.
Wertheim, E. H., Mee, V., & Paxton, S. J. (1999). Relationships among
adolescent girls’ eating behaviors and their parents’ weight-related atti-
tudes and behaviors. Sex Roles: A Journal of Research, 41, 169–187.
Wertheim, E. H., Paxton, S. J., & Blaney, S. (2004). Risk factors for body
image dissatisfaction. In J. K. Thompson (Ed.), Handbook of eating
disorders and obesity (pp. 463– 494). Hoboken, NJ: Wiley.
Wertheim, E. H., Paxton, S. J., Schutz, H. K., & Muir, S. L. (1997). Why
do adolescents girls watch their weight? An interview study examining
socio-cultural pressures to be thin. Journal of Psychosomatic Research,
42, 345–355.
Whitbourne, S. K. (2002). The aging individual: Physical and psycholog-
ical perspectives (2nd ed). New York: Springer.
Wichstrom, L. (1999). The emergence of gender difference in depressed
mood during adolescence: The role of intensified gender socialization.
Developmental Psychology, 35, 232–245.
Received April 8, 2005
Revision received December 21, 2005
Accepted December 28, 2005
899
BODY DISSATISFACTION IN ADOLESCENT GIRLS AND BOYS
... Resilience has been identified as a criterion for ED recovery [31,32], but data are still scarce. In addition, since EDs are often considered to primarily affect adolescent and young adults, studies on ED have been largely conducted among these populations [33][34][35]. However, prevalence of ED among adults is not negligible [36][37][38], which justifies studying risks and protective factors of ED in an adult population. ...
... EDs have also been correlated with dieting [11], which has been shown to be negatively associated with resilience [25,26]. Peer influence is also a known risk factor for EDs [11,12,33]. For example, in women, having a college roommate who was dieting significantly predicted drive for thinness and bulimia incidence ten years later [58]. ...
... For example, in women, having a college roommate who was dieting significantly predicted drive for thinness and bulimia incidence ten years later [58]. Friends' or parents' dieting has also been suggested as a potential predictor of body dissatisfaction [33], constant dieting [59], unhealthy/extreme weight control behavior [59] as well as binge eating in adolescents [59]. In addition, it has been suggested that individuals tend to associate with peers with similar personality [60,61]. ...
Article
Full-text available
Resilience is a positive psychological trait associated with a lower risk of some physical and mental chronic diseases and could be an important protective factor against eating disorders (EDs). The aim of this study was to assess cross-sectional and longitudinal associations between resilience and ED in a large cohort of French adults. In 2017, a total of 25,000 adults from the NutriNet-Santé cohort completed the Brief Resilience Scale (BRS). ED symptoms were measured in 2017 and 2020, with the Sick-Control-One-Fat-Food (SCOFF) questionnaire. Cross-sectional and longitudinal associations between resilience and EDs were analyzed using logistic regression, controlling for sociodemographic and lifestyle characteristics. Cross-sectional analyses showed that more resilient participants exhibited EDs less frequently than did less resilient participants (p < 0.0001). Longitudinal analyses showed that, during the three years of follow up, higher resilience was negatively associated with incident EDs (OR: 0.67, 95%CI: 0.61–0.74), persistent EDs (0.46 (0.42–0.51)), and intermittent EDs (0.66 (0.62–0.71)), compared with no ED. More resilient participants were also less likely to have a persistent ED than to recover from EDs (0.73 (0.65–0.82)). This study showed that resilience was associated with less ED symptoms and a higher chance of recovery.
... Since body image represents a large part of a person's self-concept, it is unsurprising that self-esteem and body dissatisfaction are related, [20] and high self-esteem reduces the chances of bulimia in females [21]. Research has, till now, associated low self-esteem with body dissatisfaction and psychological distress [22][23][24][25]. In fact, one study found that being female and having low self-esteem was most predictive of body image dissatisfaction for participants with healthy body weight [26]. ...
... We did not find any significant difference in the average scores of BSQ-34 in males versus females. In literature regarding the role of gender the findings have been inconsistent, as some authors [17,22,45] reports higher BID risk in females while an Indian study [46] reported higher BID in males compared to females probably because in their study male subsample had significantly higher BMI than females. In literature it appears, correlation between BMI and negative Body Image is higher in women than in men [47] maybe because most research regarding risk factors for weight gain and body image disorder has been conducted in college women, but some study reports college men are at risk too [48,49]. ...
Article
Full-text available
Background: Body Image is known to be influenced by Body Mass Index (BMI), self-esteem and personality. Body image dissatisfaction (BID) is associated with depression, anxiety, low self-esteem and maladaptive eating pattern. Therefore, it is important to identify factors that can contribute to BID. Material and Methods: This study is observational cross-sectional study on 200 medical doctors in India. Data collection was by self-reported questionnaires: Sociodemographic proforma, Body Shape Questionnaire-34 for BID, Rosenberg's Self-Esteem Scale, and Big Five Inventory for Personality traits. Results: Out of 200, 23.5% (n=47) participants had BID. BID significantly correlated with higher body mass index (BMI) (r = .402, P = 0.000) and lower self-esteem (r = −-.200, P =0.004). Among the personality traits, neuroticism (r = .309, P = 0.000) showed significant positive correlation and openness (r =-.309, P = 0.000) extraversion (r=-.139, P = 0.05) conscientiousness (r =-.188, P = 0.004) agreeableness (r =-.176, P = 0.014) had significant negative correlation. Conclusion: BMI, self-esteem, personality have a role in BID and this knowledge can be further utilized in identifying the contributing factors for development of BID and its psychiatric consequences which might serve as targets for designing intervention.
... Consistently, body dissatisfaction emerges as the most potent risk factor for EDs for women (4). Body dissatisfaction also has wide-ranging effects on health outcomes, including decreased self-esteem and engagement in physical activity, and increased risk for depression (2,5). In contrast, internalized weight stigma, also referred to as self-directed weight stigma, involves holding negative attitudes about oneself because of self-perceived excess body weight and devaluation of the self, based on societal pressures (6). ...
Article
Full-text available
Both body dissatisfaction and internalized weight stigma have been identified as risk factors for many negative health outcomes for women, including depression and eating disorders. In addition to these contributions, these concepts have been found to overlap to various degrees in existing literature. We conducted a systematic review and meta-analysis on articles published prior to February 2022 to demonstrate the conceptual and measurement overlap between body dissatisfaction and internalized weight stigma as currently quantified. We identified 48 studies examining the interrelation between body dissatisfaction and internalized weight stigma in predominantly female samples. Stronger correlations between these two constructs, some bordering on multicollinearity, were prevalent in community samples compared to clinical samples and with some but not all the commonly used measures in the body image and weight stigma fields. Body mass index (BMI) moderated these relations such that individuals with higher self-reported BMI were more likely to report lower correlations between the constructs. This concept proliferation, stronger for individuals with lower BMIs and community samples, necessitates the need change how we conceptualize and measure body dissatisfaction and internalized weight stigma. To this end, we conducted study two to refine existing measures and lessen the degree of measurement overlap between internalized weight stigma and body dissatisfaction, particularly in community samples of women. We aimed to clarify the boundaries between these two concepts, ensuring measurement error is better accounted for. Female university students completed existing measures of body satisfaction and internalized weight stigma, which were analyzed using an exploratory followed by a confirmatory factor analysis. In our attempts to modify two existing measures of internalized weight stigma and body dissatisfaction, the majority of the internalized weight stigma items were retained. In contrast, most of the body dissatisfaction items either cross-loaded onto both factors or loaded on to the internalized weight stigma factor despite being intended for the body dissatisfaction factor, suggesting that the measurement issues identified in recent prior research may be due not only to the way we conceptualize and quantify weight stigma, but also the ways in which we quantify body dissatisfaction, across the existing corpus of body dissatisfaction scales.
... Body dissatisfaction comprises the cognitive-affective component of the broader body image construct (Carey & Preston, 2019;Cornelissen et al., 2019). Defined as dislike of one's body or appearance (Fairburn, 2008), body dissatisfaction is common in women from Western cultures (Holmqvist & Frisén, 2010;Jaeger et al., 2002;Warren et al., 2005), and is a risk factor for low selfesteem, symptoms of depression, and eating disorders (Fairburn, 2008;Paxton et al., 2006). It is unsurprising therefore that researchers have sought to identify factors that contribute to body dissatisfaction. ...
Article
Objective: Although personality traits have been found to be associated with body dissatisfaction for women in the general population, little research has explored these associations for people with eating disorders. Furthermore, it is unknown whether these associations are direct or are mediated by other factors. In this cross-sectional study, secondary analyses of data from two clinical trials were conducted to determine which personality dimensions contributed to body dissatisfaction in women with bulimia nervosa, and whether low self-esteem and depression mediate these associations independently or in serial. Method: Participants were 193 women with bulimia nervosa. Participants completed self-report measures of temperament and character, body dissatisfaction, low self-esteem, and depression before receiving treatment for their eating disorder. Results: The temperament dimension, harm avoidance, contributed significantly to body dissatisfaction. Serial mediation analyses showed this association was fully mediated, revealing two significant indirect effects. The first was through low self-esteem and the second through depression and low self-esteem in serial. Conclusions: Findings suggest body dissatisfaction in women with bulimia nervosa may be indirectly targeted through addressing harm avoidance, depression and low self-esteem.
... Given that body dissatisfaction and weight bias are associated with individual health consequences (Forney & Ward, 2013;Paxton et al., 2006;Stice & Shaw, 2002), the significant decrease in scores on these constructs may correlate with improved well-being. Given the increasing demands placed on teachers, improving the well-being of teachers prior to entry into the profession may help with the increase in burnout and early attrition noted in the profession (Burke et al., 2007). ...
Article
Comprehensive school health (CSH) is a framework that can support teachers in the delivery of health-related content as well as supporting health promotion within the school community. In this study, 222 Bachelor of Education students completed surveys at the beginning and end of a mandatory six-week course on CSH with a body weight-neutral focus. Following the course, participants had significantly positively increased attitudes toward CSH, self-efficacy to teach using CSH, and weight-related attitudes. These results support CSH as a useful framework for teachers in their future practice to improve the wellness of students, teachers, and the broader school community.
... For instance, feminist models emphasise the role of society in promoting thinness as a feminine ideal (Bowden 2012;Bordo 1992;Orbach 1978), while others focus on the role that controlling or hostile family relationships play in the emergence of eating disorders (e.g. Byely et al. 2000;Legrand & Briend 2015;Paxton & Eisenberg 2006). Without wanting to dispute the role these factors play, we draw attention not to the way disordered eating practices 2 in AN arise but to a way in which they are supported and sustained by others. ...
Article
Full-text available
Anorexia Nervosa (AN) is an eating disorder characterised by self-starvation. Accounts of AN typically frame the disorder in individualistic terms: e.g., genetic predisposition, perceptual disturbances of body size and shape, experiential bodily disturbances. Without disputing the role these factors may play in developing AN, we instead draw attention to the way disordered eating practices in AN are actively supported by others. Specifically, we consider how Pro-Anorexia (ProAna) websites-which provide support and solidarity, tips, motivational content, a sense of community, and understanding to individuals with AN-help drive and maintain AN practices. We use C. Thi Nguyen's work on epistemic "echo chambers", along with Maria Lugones' work on "worlds" and "ease", to explore the dynamics of these processes. Adopting this broader temporal and intersubjective perspective, we argue, not only helps to further illuminate the experiential character of AN but also has important clinical and therapeutic significance.
... More anxiety arising from the negative judgments of others [36][37][38], showing a beneficial influence on body-image-related preoccupation [35,39]. On the other hand, individuals with low self-esteem appear to be more vulnerable to comments concerning their bodies, and more dissatisfied with physical aspects of their bodies over time [36,40]. Previous evidence suggests that increased self-esteem could play a protective role against body image concerns [41]. ...
Article
Full-text available
As with other addictions, exercise addiction can severely impact individuals’ lives and have significant psychophysical consequences. Consequently, the study of the mechanisms involved in this psychopathological condition has great clinical and practical relevance. Therefore, the goal of the present study was to explore the risk factors and protective factors for exercise addiction, with a particular focus on the roles of alexithymia, body image concerns, and self-esteem. A sample of 288 regular exercisers (mean age = 28.35 years, SD = 8.26; 72% females, 18% males) completed the Exercise Addiction Inventory, 20-Item Toronto Alexithymia Scale, Body Image Concern Inventory, and Rosenberg Self-Esteem Scale. Data were analyzed by implementing a moderated mediation model. Results showed a significant and positive association between alexithymia and exercise addiction, totally mediated by body image concerns. Furthermore, self-esteem showed a relevant moderating effect, such that at high levels of self-esteem the effect of alexithymia on body image concerns became insignificant. Such data have important implications, highlighting some core variables on which it might be useful to keep a focus in order to elaborate tailored interventions, from both preventive and treatment perspectives.
Article
Background: Despite the prevalence of body dissatisfaction among young Indonesian women and its consequential negative impacts, there are currently no evidence-based, culturally appropriate interventions to tackle this issue. Therefore, there is a need to develop scalable, cost-effective, and accessible interventions to improve body image among this population. Objective: This paper describes the study protocol of a parallel randomized controlled trial to evaluate the effectiveness of Warna-Warni Waktu, a social media-based intervention that aims to reduce state and trait body dissatisfaction and improve mood among young Indonesian women aged 15-19 years. Methods: The trial will take place online. Approximately 1800 young women from 10 cities in Indonesia, evenly split across the ages of 15-19 years, will be recruited via a local research agency's established research panel. Participants will be randomly allocated to the intervention condition or a waitlist control condition. The intervention consists of six 5-minute videos, with each video supplemented with up to five brief interactive activities. The videos (and associated activities) will be delivered at a rate of one per day across 6 days. All participants will complete three self-report assessments: at baseline (Day 1), 1 day following the intervention (Day 9), and 1 month following the intervention (Day 36). The primary outcome will be change in trait body dissatisfaction. Secondary outcomes include change in internalization of appearance ideals, trait mood, and skin shade satisfaction. Intervention effectiveness on these outcomes will be analyzed using linear mixed models by a statistician blinded to the randomized condition. Intervention participants will also complete state measures of body satisfaction and mood before and after watching each video to assess the immediate impact of each video. This secondary analysis of state measures will be conducted at the within-group level. Results: Recruitment began in October 2021, with baseline assessments underway shortly thereafter. The results of the study will be submitted for publication in 2022. Conclusions: This is the first study to evaluate an eHealth intervention aimed at reducing body dissatisfaction among young Indonesian women. If effective, the intervention will be disseminated to over half a million young women in Indonesia via Facebook, Instagram, and YouTube. Trial registration: ClinicalTrials.gov NCT05023213; https://clinicaltrials.gov/ct2/show/NCT05023213. International registered report identifier (irrid): PRR1-10.2196/33596.
Article
This study examined the temporal sequence of the relationship between social media use and body dissatisfaction in adults. A representative sample of adults (19–92 years old; M = 52.83, SD = 13.43; 62.02% women, 37.98% men) completed measures of social media use, body dissatisfaction, age, gender, BMI, and demographic variables in 2015, 2016, 2017, 2018 and 2019 (N = 6258) in the New Zealand Attitudes and Values Study. In the full sample, higher social media use was significantly associated with higher body dissatisfaction one year later, as was higher body dissatisfaction with higher social media use one year later after controlling for body dissatisfaction/social media use (T-1), gender, age, BMI, ethnicity, relationships status, and SES. Effects were small. The prospective pathway from social media use to body dissatisfaction was significant in all age groups but the reverse relationship was only significant in the middle aged and older groups. Both pathway directions were significant in women but only the pathway from body dissatisfaction to social media use was significant in men. The research has limitations and replication is required. However, findings suggest raising awareness about how to use social media positively across the broad community, not merely in adolescents, may be worthwhile.
Preprint
This study examined the temporal sequence of the relationship between social media use and body dissatisfaction in adults. A representative sample of adults (19–92 years old; M = 52.83, SD = 13.43; 62.02% women, 37.98% men) completed measures of social media use, body dissatisfaction, age, gender, BMI, and demographic variables in 2015, 2016, 2017, 2018 and 2019 (N = 6,258) in the New Zealand Attitudes and Values Study. In the full sample, higher social media use was significantly associated with higher body dissatisfaction one year later, as was higher body dissatisfaction with higher social media use one year later after controlling for body dissatisfaction/social media use (T-1), gender, age, BMI, ethnicity, relationships status, and SES. Effects were small. The prospective pathway from social media use to body dissatisfaction was significant in all age groups but the reverse relationship was only significant in the middle aged and older groups. Both pathway directions were significant in women but only the pathway from body dissatisfaction to social media use was significant in men. The research has limitations and replication is required. However, findings suggest raising awareness about how to use social media positively across the broad community, not merely in adolescents, may be worthwhile.
Article
Eating attitudes and behaviors, body image, and psychological functioning were evaluated in 98 female college students: 36 African-Americans, 34 Asian-Americans, and 28 Caucasians. African-Americans had significantly higher body mass index than either Asian-American or Caucasians. In contrast, Caucasians reported greater levels of disordered eating and dieting behaviors and attitudes and greater body dissatisfaction than did Asian-Americans and African-Americans who differed little on these measures. The nature of variability in these eating behaviors and attitudes and body image was also examined within each of the three groups. A generally consistent pattern emerged within each racial group: low self-esteem and high public self-consciousness were associated with greater levels of problematic eating behaviors and attitudes and body dissatisfaction. A history of being teased about weight and size was associated with problematic eating behaviors and attitudes and body dissatisfaction in African-Americans and Caucasians but not in Asian-Americans. The findings suggest that there exist important racial differences on various aspects of eating, dieting, and body image in college women. Contrary to hypothesis, the degree of acculturation and assimilation within the African-American and Asian-American groups was unrelated to variability in these domains. © 1995 by John Wiley & Sons, Inc.
Article
This study examined factors that influence body image and strategies to either lose weight or increase muscle among children. Participants were 237 boys and 270 girls. Body mass index (BMI), body dissatisfaction, cognitions and behaviors to both lose weight and increase muscles, as well as self-esteem and positive and negative affect, were evaluated. Self-esteem was associated with body satisfaction, positive affect predicted strategies to lose weight and increase muscles, and negative affect predicted body dissatisfaction and cognitions to lose weight and increase muscles. Boys were more likely to focus on changing muscles. Respondents with higher BMIs were more focused on losing weight but not muscle. The discussion focuses on health risk behaviors related to eating and exercise among children.
Chapter
Ever since G. Stanley Hall's (1904) seminal work a century ago, peer relationships have been regarded as a central feature of American adolescence. From the early years through the present, researchers have remained decidedly ambivalent about the effects of peers on American adolescents (Berndt, 1999), but few deny the significance of peer relationships and interactions during this stage of life. Do peers comprise a supportive social context that fosters identity and helps to socialize youth into adult roles, or do they form an arena for frivolous and delinquent activity, with patterns of interaction that undermine autonomy and self-esteem? In this chapter I overview some of the major features of peer relations that have occupied researchers' attention over the past 10 or 15 years. Insights emerging from their studies underscore the complexity of adolescent peer relations and clarify the conditions under which peer interactions foster healthy or unhealthy development.
Article
It has been said that obese persons are the last acceptable targets of discrimination.1-4 Anecdotes abound about overweight individuals being ridiculed by teachers, physicians, and complete strangers in public settings, such as supermarkets, restaurants, and shopping areas. Fat jokes and derogatory portrayals of obese people in popular media are common. Overweight people tell stories of receiving poor grades in school, being denied jobs and promotions, losing the opportunity to adopt children, and more. Some who have written on the topic insist that there is a strong and consistent pattern of discrimination, 5 but no systematic review of the scientific evidence has been done.
Article
This study explored friendship variables in relation to body image, dietary restraint, extreme weight-loss behaviors (EWLBs), and binge eating in adolescent girls. From 523 girls, 79 friendship cliques were identified using social network analysis. Participants completed questionnaires that assessed body image concerns, eating, friendship relations, and psychological, family, and media variables. Similarity was greater for within than for between friendship cliques for body image concerns, dietary restraint, and EWLBs, but not for binge eating. Cliques high in body image concerns and dieting manifested these concerns in ways consistent with a high weight/shape-preoccupied subculture. Friendship attitudes contributed significantly to the prediction of individual body image concern and eating behaviors. Use of EWLBs by friends predicted an individual's own level of use.